Recombinant Human Phenylalanine Hydroxylase: Novel Regulatory and Structural Properties

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 333, No. 1, September 1, pp. 85–95, 1996 Article No. 0367

Recombinant Human Phenylalanine Hydroxylase: Novel Regulatory and Structural Properties Devanand Kowlessur,1 Bruce A. Citron,2 and Seymour Kaufman Laboratory of Neurochemistry, National Institute of Mental Health, National Institutes of Health, Room 3D30, Building 36, Bethesda, Maryland 20892

Received May 22, 1995, and in revised form March 28, 1996

Recombinant human liver phenylalanine hydroxylase (PAH) expressed in Escherichia coli has been purified to homogeneity. The recombinant enzyme exists in solution as a mixture of 80% tetramers and 20% dimers. A study of the kinetic properties of the enzyme indicates that compared to the recombinant and the native rat liver enzymes, the recombinant human enzyme is in an activated state. This conclusion is supported by the finding that its catalytic activity is only marginally stimulated by incubation with either phenylalanine or lysolecithin. In contrast, the native and the recombinant rat liver enzymes are activated 8- to 25-fold, respectively, when preincubated with phenylalanine or lysolecithin. In the absence of activators, the ratio of the hydroxylase activity in the presence of 6-methyl-5,6,7,8-tetrahydropterin compared to the activity in the presence of (6R)-5,6,7,8-tetrahydrobiopterin (BH4), which is an index of the state of activation of the enzyme, is 4 for the human recombinant PAH compared to a value of 12 for the recombinant rat liver enzyme. Furthermore, the Km for phenylalanine in the presence of BH4 is 0.050 mM, a value that is one-fifth that of the recombinant rat liver enzyme. Covalent modification of the human enzyme by phosphorylation with protein kinase A provides further evidence that the human enzyme is in a substantially activated state. Phosphorylation, which results in the incorporation of 0.6 mol of phosphate/mol of subunit, leads to only a modest activation of 1.5-fold compared to about a 3-fold activation seen after phosphorylation of the native and the recombinant rat liver enzymes. Moreover, the recombinant human liver enzyme is less sensitive than the rat liver enzyme to stimulation by lysolecithin when tryptophan is the substrate. Just as is true for the rat liver enzyme, the apparent Km values for tryptophan and phenylalanine vary with the pterin 1

To whom correspondence should be addressed. Fax: (301) 4809284; Internet: [email protected]. 2 Present address: Neurobiology Research Lab 151, VA Medical Center, 4801 Linwood Blvd., Kansas City, MO 64128.

cofactor employed. The ability of 7-tetrahydrobiopterin (7-BH4) to substitute for the natural cofactor tetrahydrobiopterin has been studied in vitro. The apparent Km for 7-BH4 for the recombinant human enzyme is 0.2 mM and the Km for phenylalanine is 0.05 mM. The hydroxylase reaction is severely inhibited by 7-BH4 in the presence of physiological concentrations of BH4 . This inhibition can be overcome by a decrease in the concentration of phenylalanine. The implications of these novel properties of human PAH for phenylalanine homoestasis in man are discussed.

PAH3 (EC 1.14.16.1) is a non-heme iron mono-oxygenase that catalyzes the conversion of phenylalanine to tyrosine in the presence of the cofactor tetrahydrobiopterin and molecular oxygen (1). In mammals, this hydroxylation is the rate-limiting step in the complete catabolism of phenylalanine and might thus be expected to be under strict metabolic control. During the hydroxylation reaction, phenylalanine is converted to tyrosine and tetrahydrobiopterin is converted first to 4a-hydroxytetrahydrobiopterin, and then to quinonoid dihydrobiopterin; the latter reaction is catalyzed by 4a-carbinolamine dehydratase. Tetrahydrobiopterin is regenerated from quinonoid dihydrobiopterin by dihydropteridine reductase with NADH as the preferred coenzyme. In the absence of dihydropteridine reductase, the unstable quinonoid dihydrobiopterin rearranges nonenzymatically to 7,8-dihydropterin (1). Considerable interest has been focused on PAH because its deficiency leads to the genetic disease phenylketonuria. Variant forms of the disease have also been 3 Abbreviations used: BH4 , (6R)-5,6,7,8-tetrahydrobiopterin; 6MPH4 , 6-methyl-5,6,7,8-tetrahydropterin; 7-BH4 , (7R)-5,6,7,8-tetrahydrobiopterin; SDS, sodium dodecyl sulfate; PAH, phenylalanine hydroxylase; IPTG, isopropyl-b-D-thiogalactopyranoside; PKA, protein kinase A, catalytic subunit.

85

0003-9861/96

AID

ARCH 9579

/

6b1f$$$$41

07-29-96 14:43:01

arcalc

AP: Archives

86

KOWLESSUR, CITRON, AND KAUFMAN

reported that are due to genetic defects in either the biosynthesis or the turnover of tetrahydropbiopterin (2). One of the latter conditions has been traced to a deficiency of 4a-carbinolamine dehydratase and is characterized by the uninary excretion of 7-substituted biopterin (7-BH4) (3–5), a pterin that has been shown to inhibit rat liver PAH (6, 7). Inhibition of human PAH by 7-BH4 , however, has not yet been reported. PAH isolated from human liver been partially characterized (8, 9). More recently, the full-length cloned human enzyme has been expressed. Although some of its kinetic characteristics were determined, none of its regulatory properties have been reported (10). In the present report, we describe some of the structural and catalytic properties of the recombinant human enzyme that might be relevant to understanding the in vivo regulation of its activity. EXPERIMENTAL PROCEDURES Materials. Plasmid vector pET3d was obtained from Novagen. A full-length human PAH clone, TPAH/, was generously provided by Fred Ledley and Savio Woo (Baylor). Host Escherichia coli strains were DH5 from Bethesda Research Labs (Bethesda, MD) and BL21(DE3) from Novagen. Restriction endonucleases and T4 DNA ligase were obtained from Bethesda Research Labs and New England Biolabs (Beverly, MA). Dideoxy sequencing reagents were from United States Biochemicals and were used according to the manufacturer’s recommendations. Reagents for the polymerase chain reaction were obtained from Perkin–Elmer (Norwalk, CT). Bacterial growth medium was LB (0.5% NaCl, 1% tryptone, 0.5% yeast extract) supplemented with 100 mg/ml ampicillin, 0.1 mM ferrous ammonium sulfate, and 0.5 mM isopropyl-b-D-thiogalactopyranoside (IPTG) (for induction). Phenylalanine; lysolecithin; glucose 6-phosphate, tryptophan, glucose-6-phosphate dehydrogenase (from Leuconostoc mesenteroides), dihydropteridine reductase, and protein kinase A, catalytic subunit, were obtained from Sigma. (6R)-Tetrahydrobiopterin, 7-tetrahydrobiopterin, and 6-methyltetrahydropterin were obtained from Dr. B. Shirck’s laboratories, Jona, Switzerland. [g-32P]ATP was purchased from New England Nuclear Corp. (Boston, MA). Catalase was obtained from Boehringer Mannheim. DNA manipulations. DNA amplification of the human PAH coding sequence was performed in a 100-ml reaction with 100 ng/ml plasmid TPAH/ (10) as a template and 0.5 mM oligo O178 (5* ACGGCTAACT TCATGAGCAC TGCGGTCCTG GAAAACC) as the 5* primer and 0.5 mM O146 (5* CTGAACTAGC TGATCAACAG ATTCACAGCT GACAG) as the 3* primer, 2.5 mM MgCl2 , 200 mM each deoxynucleotide triphosphate, and 25 U/ml Taq polymerase (AmpliTaq). The amplification was conducted with an initial denaturation at 987C for 10 min with all components except polymerase, then, after addition of the enzyme at 907C, 10 cycles of denaturation at 957C for 1 min, annealing at 607C for 1 min, and extension at 727C for 1 min. The 5* primer was designed with a BspHI restriction site and the 3* primer contained a BclI site to facilitate substitution of the human PAH coding sequence for the small NcoI–BamHI fragment of the vector, pET3d. Induction of the human PAH was performed as previously described for the recombinant rat enzyme (11). PAH. PAH from rat liver was purified as previously described (12). Recombinant human PAH was purified by a modification of this method (11). The amino acid sequence at the N-terminal region of the purified human PAH was determined using a solution (10 ml) of

AID

ARCH 9579

/

6b1f$$$$42

07-29-96 14:43:01

2 mg/ml enzyme diluted with 200 ml of water and applied to a Prospin device (Applied Biosystem). The resulting polyvinylidene difluoride blot was washed with 20% methyl alcohol prior to sequencing in the blot cartridge on a Model 447A pulsed-liquid protein sequencer equipped with a Model 120A PTH analyzer (Applied Biosystem) using methods and cycles supplied by the manufacturer. Data were collected and analyzed on a Model 610A data analysis system (Applied Biosystem). PAH activity was monitored by either of two methods. In the first method, the phenylalanine-dependent oxidation of NADH was monitored at 340 nm (12). Enzyme assays were performed in 0.1 M phosphate buffer, pH 6.8, at 257C at atmospheric oxygen tension. The assay mixture contained 0.1 mg/ml catalase, 125 mM NADH, and an excess of dihydropteridine reductase. The concentrations of the enzyme, phenylalanine, and pterin cofactor were varied for each individual experiment. The enzymes were incubated for 3 min in the presence of phenylalanine before starting the reaction. An extinction coefficient of 6220 M01 cm01 was used to calculate the amount of NADH consumed. The initial rates were corrected for the contribution due to autooxidation of the tetrahydrobiopterin. During the purification procedure of the enzyme, a second method of assaying PAH was used that is based on the colorimetric determination of the amount of tyrosine formed. Protein fractions at various stages of the purification procedure were assayed for PAH activity at 257C for 30 min in the presence of 6MPH4 as cofactor. The complete reaction mixture contained the following components in a final volume of 1 ml: 0.1 M potassium phosphate buffer, pH 6.8, 5 mM phenylalanine, 0.1 mg/ml catalase, 0.35 mM NADH, 10 mM glucose 6-phosphate, and excess dihydropteridine reductase and glucose-6-phosphate dehydrogenase. The reactions were started with addition of 0.5 mM tetrahydropterin. After 30-min incubation at atmospheric oxygen tension, the reaction was stopped by addition of 12% trichloroacetic acid, and the tyrosine formed was determined colorimetrically by the nitrosonaphthol method (12). Protein concentrations were determined by the method of Lowry with bovine serum albumin as standard (13). Polyacrylamide gel electrophoresis in the presence of SDS was performed according to Laemmli (14). High-pressure gel filtration. HPLC was carried out essentially according to the procedure described by Horiike et al. (15) under nondenaturing conditions with the use of a Gilson Model 305 highpressure liquid chromatography system monitoring protein by absorbance at 280 nm. The gel-filtration column was a TSK-GEL (7.5 1 600 mm; Altex), equilibrated with 0.1 M potassium phosphate buffer, pH 6.8. Protein samples (1–4 mg/ml) were applied in a volume of 25–50 ml at a flow rate of 1 ml/min at 257C. The column was calibrated with the following protein standards, supplied by Pharmacia: catalase (Mr 232,000), aldolase (Mr 158,000), ferritin (Mr 440,000), thyroglobulin (Mr 669,000), ovalbumin (Mr 43,000), chymotrypsinogen A (Mr 25,000), ribonuclease A (Mr 13700), and Blue 2000 Dextran. Phosphorylation. The recombinant human PAH was phosphorylated in a volume of 0.2 ml at 307C, in a reaction mixture containing 10 mM Tris HCl, pH 7.4, 1 mM MgCl2 , 0.8 mM ATP, 5 mg of protein kinase catalytic subunit, and 400 mg of recombinant PAH. From the reaction mixture, 10-ml aliquots were withdrawn at 10-min intervals to measure activity. In experiments to measure the incorporation of 32 Pi from [g-32P]ATP into the recombinant enzyme, the conditions used for incubation were similar, except that ATP was substituted by [g-32P]ATP; 20-ml aliquots were withdrawn and the reaction was stopped at various time intervals by the addition of 100% trichloroacetic acid, and the resulting precipitate was collected by centrifugation. After removal of the supernatant, the pellet was dissolved in 0.1 M NaOH and precipitated immediately again with trichloroacetic acid. The pellets were treated three times with 0.1 M NaOH and trichloroacetic acid; the final pellets were dissolved in 0.1 M NaOH and the incorporation of radioactivity was determined.

RESULTS

Construction of human PAH expression plasmid. To provide large quantities of enzyme for quantitative

arcalc

AP: Archives

87

HUMAN PHENYLALANINE HYDROXYLASE TABLE I

Purification of Recombinant Human PAH Total protein (mg)

Procedure 1. 2. 3. 4. 5.

Extract Ammonium sulphate precipitation Phenyl–Sepharose column DEAE Ultrafiltration

ARCH 9579

/

6b1f$$$$42

Specific activity (mmol/min/mg)

Yield (%)

75 58 17 6.0 7.3

0.03 0.19 0.85 1.3 2.6

100 77 23 8 9.8

2520 313 20 4.5 2.8

analyses, and to facilitate future genetic manipulations, we placed the human PAH coding sequence in the pET bacterial expression system. Polymerase chain amplification was performed with full-length human PAH clone TPAH/ (10) as a template and priming oligonucleotides to produce a DNA fragment with restriction sites near the termini for insertion into the efficient expression vector pET3d. The amplified band and vector were each digested with restriction enzymes that produced compatible, cohesive ends, ligated together, and inserted into E. coli host cells by transformation. The composite DNA molecules were isolated and purified and the absence of any misincorporations during the DNA manipulations was confirmed by complete DNA sequence analysis. Finally, to allow expression of human PAH, host cells harboring an inducible T7 RNA polymerase were transformed with pLNCB111, which contains the wild-type human PAH coding sequence in the expression vector pET3d. Expression of the human PAH in E. coli was induced with IPTG for 6 h during log-phase growth. The cells were harvested by centrifugation and stored at 0707C until they were disrupted by sonication prior to isolation of the hydroxylase. The procedure for the purification of recombinant human PAH is summarized in Table I. After DEAE-cellulose ion-exchange chromatography of the enzyme and concentration by Amicon-P30 filtration, the final protein was estimated to be 90% pure by analysis on SDS– polyacrylamide gel and staining with Coomassie blue. The molecular weight of the cloned enzyme (monomer) was determined to be 52,000, a value that is the same as that reported for the native liver enzyme (16). The following N-terminal amino acid sequence for the first 20 residues was observed: STAVLENPGLGRKLSDFGQE. The observed sequence is identical to that predicted by the cDNA. No other significant minor sequences were observed. When assayed with 6MPH4 , the recombinant enzyme has a specific activity of 2.6 mmol/min/mg. This value is about one-sixth that of rat liver recombinant PAH (30) but is similar to values obtained for the human enzyme isolated from liver samples (10). The oligomeric composition of the recombinant en-

AID

Total enzyme (mmol/min)

07-29-96 14:43:01

zyme was determined by high-pressure gel filtration on an HPLC column as described under Experimental Procedures. The molecular weights of the enzymes were calculated from a log plot of molecular weights against retention time. The molecular weights of the two major forms of the enzymes were estimated to be 200,000 { 5000 and 100,000 { 5000, corresponding to tetramers and dimers with retention times of 14.5 and 16.5 min (data not shown). The result indicated that the enzyme consists of a mixture of 80% tetramers and 20% dimers. Thus, the oligomeric composition of the recombinant human liver PAH is similar to that of the rat liver recombinant PAH and the enzyme isolated from rat liver (17). Kinetic properties of recombinant human and rat PAH in the presence of BH4 and 6MPH4 . A comparative kinetic study was carried out to investigate TABLE II

Michaelis–Menten Parameters for Phenylalanine Hydroxylation by Recombinant Human and Rat PAH with BH4 , 6MPH4 , and 7-BH4 as Cofactors

Cofactor (6R)BH4 (6R)BH4a 6MPH4 (7R)BH4

Km (cofactor) (mM)

Km (phenylalanine) (mM)

3 (S) (2.5) 25 (S) (23) 45 (H) (50) 200 (H) (210)

50 (H) (250) 49 (H) (260) 43 (H) (45) 55 (S)b (50)

Note. The apparent Km values were determined as described under Experimental Procedures. All experimental data were processed by computer fit to the Michaelis–Menten equation. Where sigmoidal kinetics were observed, the apparent Km shown are phenylalanine concentrations at which half-maximal rates were obtained. (H, hyperbolic; S, sigmoidal.) The parameters in parentheses refer to the recombinant rat liver hydroxylase. Each experiment was carried out in triplicate. a Phenylalanine-activated. b Km was determined from S0.5 (the half-maximal substrate concentration) where calculation of a true Km was prevented from substrate inhibition.

arcalc

AP: Archives

88

KOWLESSUR, CITRON, AND KAUFMAN

FIG. 1. Dependence of initial velocity of the PAH-catalyzed reaction on the concentration of phenylalanine with (6R)BH4 as the cofactor. The concentration of (6R)BH4 was 0.011 mM and the recombinant human PAH was present at 22.5 mg/ml. The assays were performed spectrophotometrically by monitoring the oxidation of NADH at 340 nm. The saturation curves were generated by preincubation of the enzyme at 257C for 3 min in the presence (l) or absence (s) of phenylalanine.

whether the recombinant human PAH shares similar properties with the rat liver enzyme. The values for the apparent Km for phenylalanine in the presence of 6MPH4 , as well as the values for BH4 , shown in Table II, are comparable to those of the recombinant rat enzyme and the native rat liver PAH. By contrast, the value of 0.05 mM for the apparent Km for phenylalanine calculated from the data in Fig. 1 is about one-fifth that of the recombinant rat liver enzyme and the native rat liver enzyme (18). Interestingly, this value is about the same as the normal human phenylalanine plasma level (2). The Hill coefficient is 1.1 for the recombinant human liver enzyme which is approximately 50% that found for the rat liver enzyme and 65% of the value found for rat recombinant enzyme. Preincubation with phenylalanine did not change the Km for phenylalanine (Km Å 0.049 mM) when measured with BH4 . The Km for BH4 , however, was increased from 3 to 25 mM (see Table II). The Hill coefficient was calculated to be 1.3, which is slightly higher than the value for the nonpreincubated enzyme. The effect of preincubation with phenylalanine on Vmax will be discussed below. Kinetic properties of recombinant human and rat liver PAHs with 7-BH4 . As found previously for the rat liver PAH (19), in the presence of 7-BH4 , the initial velocity versus phenylalanine concentration curve showed substrate inhibition at high concentrations and cooperativity at low substrate concentrations (Fig. 2a). Figure 2b illustrates positive cooperativity at low concentrations of phenylalanine with 7-BH4 as the cofac-

AID

ARCH 9579

/

6b1f$$$$43

07-29-96 14:43:01

tor. A comparison of these results with those obtained with BH4 (Fig. 1) indicates that the sigmoidicity is more pronounced in the presence 7-BH4 . The apparent Hill coefficient with 7-BH4 is about 2.0. In the presence of 7-BH4 the concentration of phenylalanine required to reach half-maximal activity is 50 mM (Fig. 2a). Substrate inhibition, however, prevented a determination of a reliable Km value. The apparent Km of phenylalanine is 50 mM, the same as the Km in the presence of BH4 . The dependence of the initial velocity on 7-BH4 concentration for the recombinant human PAH followed classical Michaelis–Menten kinetics with an apparent Km value 200 mM in the presence of 0.5 mM phenylalanine (Fig. 2b). This value is approximately 60-fold higher than the apparent Km for BH4 determined under comparable experimental conditions. The apparent Km values obtained for phenylalanine and 7BH4 with the recombinant human enzyme are within the same range as those of the recombinant rat liver and the native rat liver enzyme (19). Inhibition of the recombinant human liver phenylalanine hydroxylase by 7-BH4 . Although 7-BH4 has cofactor activity with recombinant human liver PAH, this pterin also acts as a potent inhibitor when BH4 is used as the cofactor. Figure 3 shows the effect of increasing the concentration of 7-BH4 at various fixed concentrations of phenylalanine. At a concentration 0.5 mM phenylalanine and 8.6 mM BH4 , the recombinant enzyme shows 50% inhibition at 1 mM 7-BH4 (Fig. 3). Moreover, at 0.1 mM phenylalanine, the enzyme is inhibited by 50% at 4 mM 7-BH4 . Interestingly, when the concentra-

arcalc

AP: Archives

HUMAN PHENYLALANINE HYDROXYLASE

89

FIG. 2. Dependence of the initial velocity of the recombinant human liver PAH-catalyzed reaction on the concentration of phenylalanine with 7-BH4 as the cofactor. Phenylalanine-dependent oxidation of NADH was followed spectrophotometrically at 340 nm. The rates are expressed in mmol of NADH oxidized per minute per milligram of protein. The concentration of 7-BH4 was 1 mM and PAH was present at 22.5 mg/ml. The full range is depicted (a) as well as an expanded graph (b) for the low substrate concentrations.

tion of phenylalanine was near the physiological concentration, i.e., 50 mM, the degree of inhibition is proportional to the concentration of 7-BH4 . These results

AID

ARCH 9579

/

6b1f$$$$43

07-29-96 14:43:01

indicate that the degree of inhibition is dependent on the concentrations of both 7-BH4 and phenylalanine, i.e., it exhibits a synergistic effect. A Ki value of 1.5 mM

arcalc

AP: Archives

90

KOWLESSUR, CITRON, AND KAUFMAN

FIG. 3. Effect of increasing phenylalanine concentration at fixed concentration of BH4 on the inhibition of recombinant human liver PAH by 7-BH4 . The assays were monitored at 340 nm for the oxidation of NADH. The concentration of 7-BH4 was varied between 1 and 35 mM in samples containing 500 mM phenylalanine (curve a), 100 mM phenylalanine (curve b), and 50 mM phenylalanine (curve c). The data are presented as percentage of the remaining activity in the presence and absence of 7-BH4 . The concentration of BH4 was 8.6 mM.

was obtained for 7-BH4 , compared to a Km of 200 mM when this pterin serves as cofactor (data not shown). The results shown in Fig. 3 are reminiscent of the results obtained when the effect of 7-BH4 on rat liver PAH was studied (19). The stoichiometry of phosphorylation and its influence on the catalytic properties of the recombinant human PAH. Covalent modification of the human recombinant PAH by phosphorylation mediated by the catalytic subunit of protein kinase was carried out to investigate whether the recombinant human enzyme is activated under these conditions. Treatment of the enzyme with the catalytic subunit of protein kinase A and [g-32P]ATP led to the incorporation of 0.6 mol of [32Pi]phosphate per mole of hydroxylase subunit. Phosphorylation leads to a 1.5-fold stimulation of PAH activity in Vmax when assayed in the presence of BH4 (Table III). In contrast, phosphorylation of both the recombinant and the native rat liver enzyme (20–22) resulted in a 2- to 4-fold stimulation of activity when assayed in the presence of BH4 . It should be noted that Martinez et al. (23) also reported only a modest (1.2fold) increase in activity of recombinant human PAH when the activity was assayed with BH4 . The effects of lysolecithin on the kinetic properties of the recombinant human liver PAH. Rat liver PAH catalyzes the hydroxylation of phenylalanine at a relatively slow rate in the presence of BH4 in the assay system used. This rate can be increased 15- to 30-fold by incubating the enzyme with phenylalanine or with certain phospholipids such as lysolecithin (1). Recombinant rat liver PAH shares these properties (11). In con-

AID

ARCH 9579

/

6b1f$$$$43

07-29-96 14:43:01

trast to the enzyme’s behavior in the presence of BH4 , when 6MPH4 or 7-BH4 is used as the coenzyme, catalysis is rapid with or without preactivation with phenylalanine or lysolecithin. Furthermore, the ratio of the activity in the presence of 6MPH4 to the rate with BH4 is an index of the state of activation of the enzyme, with a higher ratio indicating a low state of activation (24). For the unactivated recombinant human enzyme the ratio of hydroxylase activity in the presence of 6MPH4 and BH4 is about 4 compared to that of the rat liver recombinant enzyme which is 12 (Table III). This difference strongly supports the conclusion that human PAH is in an activated state compared to the rat liver hydroxylase. Preincubation in the presence of BH4 inhibits the activation by phenylalanine by 16% compared to the unpreincubated enzyme. As can be seen in Table III, when BH4 is used as the coenzyme, the rate of hydroxylation of phenylalanine catalyzed by recombinant human liver PAH is increased only by about 2-fold when the enzyme is activated by either phenylalanine or lysolecithin. In fact, this degree of activation is only slightly greater than that seen in the presence of 6MPH4 or 7-BH4 (Table III). The modest activation of the human enzyme by preincubation with phenylalanine or lysolecithin also indicates that it is already in a substantially activated state. Additional evidence supporting the conclusion that the human enzyme is in an activated state was obtained in the studies of the ability of recombinant human PAH to catalyze the hydroxylation of tryptophan. It has been shown that without some kind of activation, the tryptophan-hydroxylating activity of rat liver PAH in the presence of BH4 is very low and that lysolecithin stimulates this activity even more than it does the hydroxylation of phenylalanine (1). This result indicated that the ability to catalyze the hydroxylation of alternate substrates such as tryptophan is one of the characteristics of the activated enzyme. Furthermore, with the hydroxylation of tryptophan, a major part of the effect of lysolecithin is due to a decrease in the Km of tryptophan by almost 2 orders of magnitude (1). In striking contrast to these results, lysolecithin was found to stimulate only modestly the tryptophan-hydroxylating activity of human PAH in the presence of BH4 (Fig. 4). Furthermore, the apparent Km for tryptophan is essentially the same in the presence (0.46 mM) and in the absence (0.52 mM) of lysolecithin (Fig. 4). These results, therefore, provide powerful additional support for the conclusion that recombinant human PAH is in a highly activated state. To explore the possible relationship between some of the unique catalytic properties of human PAH and phenylalanine homoeostasis in vivo, it would be useful to know how the rate of phenylalanine hydroxylation varies with the concentration of phenylalanine under condition that mimic those that occur in vivo. Studies

arcalc

AP: Archives

91

HUMAN PHENYLALANINE HYDROXYLASE TABLE III

Activation of Recombinant Human and Rat Liver PAH by Phenylalanine, Lysolecithin, and Phosphorylation in the Presence of Different Pterin Cofactors Fold activation Coenzyme

Activity (mmol/min/mg)

Activator

BH4

None Phenylalanine Lysolecithin Phosphorylation None Phenylalanine Lysolecithin None Phenylalanine Lysolecithin

6MPH4

7-BH4

0.74 1.66 1.6 1.11 2.92 3.84 3.15 1.28 1.48 1.62

(0.42) (3.36) (10.50) (1.26) (5.2) (6.0) (6.8) (2.4) (2.8) (3.0)

Human PAH — 2.2 2.2 1.5 — 1.8 1.1 — 1.2 1.3

Rat PAH

8.0 25.0 3.0 1.2 1.3 1.2 1.3

Note. The reactions were monitored spectrophotometrically for the oxidation of NADH as described under Experimental Procedures. All the reactions were carried out at 257C and contained 0.1 M potassium phosphate buffer, pH 6.8, 50 mg/ml catalase, 130 mM NADH, and an excess of dihydropteridine reductase. The enzyme was preincubated at 257C for 3 min with 1 mM lysolecithin. The effect of phenylalanine on the recombinant enzyme was carried out by preincubation with 1 mM phenylalanine at 257C for 5 min prior to assay. When assaying with BH4 , the final concentration of phenylalanine was 0.25 mM. For reactions containing 6MPH4 , 5 mM phenylalanine was added to the reaction mixture. When 7-BH4 was the cofactor, 0.5 mM phenylalanine was included. The concentrations of the cofactors BH4 , 6MPH4 , and 7-BH4 were 0.040, 0.30, and 0.90 mM, respectively. The concentrations of the recombinant human and the rat enzymes were 22.0 and 2.3 mg/ml for preincubated and nonpreincubated enzymes. The results shown in parentheses are related to the recombinant rat liver hydroxylase. The activity measurements were carried out in triplicate and the standard deviation was about 10% of the measured value.

with rat liver PAH indicate that one of the key determinants of the activity of the hydroxylase is its activation by phenylalanine and the attenuation of this effect by

BH4 (25, 26) (Fig. 5). BH4 is also known to antagonize the phosphorylation-mediated activation of the hydroxylase (27, 28).

FIG. 4. Dependence of the initial velocity of the recombinant human liver PAH-catalyzed reaction on the concentration of tryptophan with BH4 as the cofactor in the presence and absence of lysolecithin. The assays were performed by following the formation of 5-hydroxytryptophan fluorometrically (12). The concentration of tetrahydrobiopterin was 40 mM and the concentration of the hydroxylase was 0.040 mg/ml. The total reaction was 0.1 ml and the incubation was carried out at 257C for 30 min.

AID

ARCH 9579

/

6b1f$$$$43

07-29-96 14:43:01

arcalc

AP: Archives

92

KOWLESSUR, CITRON, AND KAUFMAN

FIG. 5. Dependence of the rate of reaction of recombinant rat PAH on phenylalanine concentration. (Inset) Dependence of the rate of reaction on phenylalanine at low concentrations of phenylalanine. The concentration of BH4 in the reaction was 7.8 mM. The reaction was started with the addition of the hydroxylase. Hydroxylase activities were monitored at 340 nm for the oxidation of NADH.

It was therefore of interest to repeat the experiment depicted in Fig. 1, in which the reaction was initiated by the addition of BH4 (i.e., the hydroxylase was not exposed to phenylalanine prior to the addition of BH4). The rate profile under these conditions is shown in Fig. 6. As can be seen, with enzyme as the last addition the half-maximum response occurs at Ç0.25 mM phenylalanine compared to the value of Ç0.05 mM phenylalanine under the condition shown in Fig. 1, curve B. These results indicate that just as with rat liver PAH the opposing effects of BH4 and phenylalanine are also factors that can determine the state of activation of human PAH. DISCUSSION

The initial characterization of cloned human PAH purified from E. coli reported several of the catalytic properties of the enzyme (10). From these early results, the hydroxylase from this source appeared to be essentially indistinguishable from the well-characterized rat liver enzyme. In this initial study, however, none of the physical or regulatory properties of the human enzyme were explored. In this earlier study, human PAH expressed in E. coli using the pET system was found to have been subjected to limited proteolysis by host cell proteases (23). By contrast, in the present study, there was no indication that the expressed human PAH had been proteolyzed. This conclusion is based not only on the finding that the Mr of the isolated PAH corresponds to that of the intact enzyme (23), but more important,

AID

ARCH 9579

/

6b1f$$$$43

07-29-96 14:43:01

on an analysis of the N-terminal amino acid sequence. This is an important point since it is known that proteolytic removal of N-terminal region of PAH leads to activation of the enzyme (1). Although the reason for different results with the pET system are not known with certainty, it should be noted that the induction period with IPTG used in the present studies was 6 h compared to the 18- to 24-h period used by Martinez et al. (23). It is possible that our use of a much shorter induction period circumvented this problem. In the present study, we have examined in greater detail some of the characteristics of the human enzyme, especially those catalytic properties that might be relevant to the physiological regulation of the enzyme. Unexpectedly, our results indicate that the human enzyme is in a substantially activated state. This conclusion is supported by the finding that its phenylalaninehydroxylating activity is only modestly (about 2-fold) enhanced by preincubation with either lysolecithin or phenylalanine (Table III), compared to the 20- to 30fold stimulation of the rat liver enzyme and the 8- to 25-fold stimulation of the recombinant rat liver enzyme (1) (Table III) by these activators. Also supporting this conclusion is the finding that lysolecithin only doubles the tryptophan-hydroxylating activity of the human enzyme and does this without decreasing the Km for tryptophan (Fig. 4). By contrast, lysolecithin decreases the Km for the rat liver enzyme for tryptophan by almost 2 orders of magnitude (1). Another indication that the human enzyme is already substantially activated

arcalc

AP: Archives

HUMAN PHENYLALANINE HYDROXYLASE

93

FIG. 6. Dependence of the rate of reaction of the recombinant human liver PAH on phenylalanine concentration. (Inset) Dependence of the rate of reaction on phenylalanine at low concentrations of phenylalanine. The reactions were measured at 340 nm for the oxidation of NADH. The reactions were started with the addition of the human hydroxylase last.

is that phosphorylation catalyzed by protein kinase A, which leads to the incorporation of 0.6 mol of phosphate/mol of subunit, only activates about 1.5-fold, compared to the 2- to 4-fold phosphorylation-mediated activation of the rat liver enzyme and the 2.5-fold activation of the recombinant rat enzyme (11). In contrast to the results of Abita et al. (9), who failed to detect either protein kinaseA-mediated phosphorylation or activation of the human liver hydroxylase, our present results are more in line with the findings of Smith et al. (29), who reported that the incorporation of as much as 0.67 mol of phosphate per mole of hydroxylase was not accompanied by any increase in the hydroxylase activity. More recently, it has been reported that PKA catalyzes the incorporation of up to 0.97 mol of phospate per mole of human recombinant PAH, resulting in a 1.2-fold stimulation of hydroxylase activity (23). It should be noted that the present finding of only modest activation of human PAH by phosphorylation is compatible with the failure to detect any phosphorylation-mediated activation of the human liver hydroxylase in crude extracts (30). As discussed previously (1), this lack of any activation may reflect the fact that the isolated human liver enzyme contains sufficient endogenous phosphate to slightly activate it. It should be noted, however, that a high level of endogenous phosphate is an unlikely explanation for our present findings because we are unaware of other mammalian enzymes that are phosphorylated in E. coli. Another characteristic of the human enzyme that has

AID

ARCH 9579

/

6b1f$$$$43

07-29-96 14:43:01

yielded conflicting results is its molecular weight. The reported values range from 150 to 165 kDa, corresponding to a trimer (9, 31), to extreme values of 108 (32) and 275 kDa (33). By contrast, the present results indicate that both the size and the oligomeric composition of the cloned human enzyme are very similar to these values for the rat liver hydroxylase, i.e., the enzyme in solution is a mixture of 80% tetramers (Mr 200 kDa) and 20% dimers (Mr 100,000 kDa). Martinez et al. (23) have also reported that recombinant human PAH exists predominantly as dimers and tetramers. Properties of human PAH that are relevant to its physiological regulation are the Km values for its pterin coenzyme and for its amino acid substrate. The Km for BH4 of 3 mM (Table II) agrees with our previous value obtained with human PAH purified from an autopsy sample (8) and is probably not significantly different from the value of 2 mM reported for the rat liver enzyme (34). It is, however, markedly different from the value of 16 mM previously reported for the cloned human enzyme (10), a value that is more characteristic of the rat liver hydroxylase after activation by preincubation with lysolecithin (35) or phenylalanine (25). In the present study, preincubation of PAH with phenylalanine increased the Km for BH4 from 3 to 25 mM (Table II). The latter value is coherent with the value of 29 mM reported by Martinez et al. (23) who routinely preincubated their PAH samples with phenylalanine prior to assay. This procedural variable may also explain some of the other differences

arcalc

AP: Archives

94

KOWLESSUR, CITRON, AND KAUFMAN

between the Km values reported by Martinez et al. and the present ones. The present value of 50 mM for the apparent Km of phenylalanine when assayed with BH4 (Table II) agrees with those that have been published previously, i.e., 40 (8) and 50–75 mM (10). As noted previously (8), a value of 50 mM is close to normal human plasma phenylalanine levels. Together with the data showing that the human enzyme is in a substantially activated state, the low Km value for phenylalanine in the presence of BH4 , which is about one-fifth the value for the rat liver enzyme (Table II), is one of the more striking differences between the human and the rat liver enzymes. These results have important implications for our understanding of phenylalanine homeostasis in man. Previous discussions of this topic have been based on the reasonable assumption that the regulatory properties of human liver PAH would prove to be essentially the same as those of the rat liver enzyme. The hallmark of the latter enzyme is that at resting blood and hepatic levels of phenylalanine, the enzyme has very low activity but that it is poised to be massively activated in response to physiological demand, i.e., to increased levels of phenylalanine, which not only can directly activate the enzyme in vivo, but can also work synergistically with phosphorylation to further increase the enzyme’s activity (1, 36). Our present findings that human PAH is in a relatively activated state indicate that in response to a physiological bolus of phenylalanine—e.g., to a typical meal (hamburger and milkshake)—blood levels of phenylalanine would probably increase by 1.6-fold (37) and the rate of hydroxylation by 1.8-fold (Fig. 6). In contrast to this response of the human enzyme, with rat PAH a 60% increase in concentration of phenylalanine would be expected to increase the hydroxylase activity about 2.2-fold (see Fig. 5). The difference in response of the two enzymes becomes greater with greater increases in blood phenylalanine levels. Thus, a 3-fold increase in the concentration of phenylalanine (from 0.05 to 0.15 mM) (Fig. 6) would be expected to lead to a 3.6-fold increase in the hydroxylase activity of the human enzyme and a 5.3-fold increase in activity of rat PAH (Fig. 5). These comparisons indicate that a significant error would be introduced by acceptance of the prevalent assumption that the regulatory properties of the two hydroxylases are essentially the same. It should be noted that the above comparison does not take into account the fact that rat liver enzyme is activated to a greater extent than the human enzyme by PKA-mediated phosphorylation (see Table III). Since it is unlikely that increased levels of phenylalanine would lead to a comparable phosphorylation-mediated activation of human PAH, this secondary activating effect of phenylalanine would exacerbate any error that would be introduced by an unqualified extrapolation of the regulatory properties of the rat enzyme to its human counterpart.

AID

ARCH 9579

/

6b1f$$$$43

07-29-96 14:43:01

The present evidence that human hepatic enzyme is in an activated state is coherent with the conclusion reached from a limited study of some properties of PAH in a human liver biopsy sample. In that study, the 6MPH4/BH4 ratio of hydroxylase activity for the human enzyme was found to be 9 compared to a value of 44 for the rat liver enzyme, both measured in crude extracts (30). Based on this finding, it was predicted that the human hydroxylase might not be activated by phosphorylation. It is of interest that in the present study, the 6MPH4/BH4 ratio of activities, calculated from the data in Table III, is about 4, an indication that the pure cloned human enzyme may be even somewhat more activated than the human enzyme appears to be in crude liver extracts. A comparison of the 6MPH4/BH4 ratio in crude and pure PAH preparations may not be valid, however, since it is known that this ratio for the rat liver enzyme increases during the course of its purification (24). Finally, the present findings show that, just as with rat liver PAH (6, 19), 7-BH4 has cofactor activity with the human enzyme. Perhaps of greater physiological significance, however, our present results show that the human enzyme also shares with rat liver PAH its sensitivity to inhibition by micromolar concentrations of 7-BH4 and that the inhibition is exacerbated by elevated levels of phenylalanine (Fig. 3). These results with 7-BH4 provide added support for the proposal (6, 19) that the hyperphenylalaninemia that is seen in patients who excrete increased amounts of 7-BH4 is due to the inhibition of PAH by this isomer of BH4 . ACKNOWLEDGMENTS We thank Savio Woo and Fred Ledley for providing a human PAH cDNA clone; John Giovanelli for preliminary results and useful discussion; Kun Park, Margot Gibson, and Cynthia Falke for technical assistance; and Joseph Shiloach for large-scale fermentations. We also thank Howard Jaffe for sequencing the N-terminal region of the recombinant human hydroxylase.

REFERENCES 1. Kaufman, S. (1993) in Advances in Enzymology (Meister, A., Ed.), pp. 77–264, Wiley, New York. 2. Scriver, C. R., Eisensmith, R. C., Kaufman, S., and Woo, S. L. C. (1995) in The Metabolic Basis of Inherited Disease Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D., Eds.), pp. 1015–1076, McGraw–Hill, New York. 3. Dhondt, J., Guibaud, P., Rolland, M., Dorche, C., Andre, S. K., Forzy, G., and Hayte, J. (1988) Eur. J. Pediatr. 147, 153. 4. Blau, N., Curtius, H.-C., Kuster, T., Matasovic, A., Schoedon, G., Dhondt, J. L., Guibaud, P., Giudici, T., and Blaskovics, M. (1989) J. Inherited Metab. Dis. 12, 335–338. 5. Curtius, H.-C., Adler, C., Rebrin, I., Heizmann, C., and Ghisla, S. (1990) Biochem. Biophys. Res. Commun. 172, 1060–1066. 6. Adler, C., Ghisla, S., Rebrin, I., Heizmann, C. W., Blau, N., and Curtius, H.-C. (1992) J. Inherited Metab. Dis. 15, 405–408. 7. Davis, M. D., Kaufman, S., and Milstien, S. (1991) Proc. Natl. Acad. Sci. USA 88, 385–389.

arcalc

AP: Archives

HUMAN PHENYLALANINE HYDROXYLASE 8. Friedman, P. A., and Kaufman, S. (1973) Biochim. Biophys. Acta 293, 56–61. 9. Abita, J.-P., Blandin-Savoja, F., and Rey, F. (1983) Biochem. Int. 7, 727–737. 10. Ledley, F. D., Grenett, H. E., and Woo, S. L. C. (1987) J. Biol. Chem. 262, 2228–2233. 11. Citron, B. A., Davis, M. D., and Kaufman, S. (1992) Protein Expression Purif. 3, 93–100. 12. Kaufman, S. (1987) in Methods in Enzymology (Kaufman, S., Ed.), Vol. 142, pp. 3–17, Academic Press, Orlando. 13. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265–275. 14. Laemmli, U. K. (1970) Nature 227, 680–685. 15. Horiike, K., Tojo, H., Iwaki, M., Yamano, T., and Nozaki, M. (1982) Biochem. Int. 4, 477–483. 16. Iwaki, M., Parniak, M. A., and Kaufman, S. (1985) Biochem. Biophys. Res. Commun. 126, 922–932. 17. Parniak, M., and Kaufman, S. (1985) Biochemistry 24, 3379A. 18. Parniak, M. A., Davis, M. D., and Kaufman, S. (1988) J. Biol. Chem. 263, 1223–1230. 19. Davis, M. D., Ribeiro, P., Tipper, J., and Kaufman, S. (1992) Proc. Natl. Acad. Sci. USA 89, 10109–10113. 20. Abita, J. P., Milstien, S., Chang, N., and Kaufman, S. (1976) J. Biol. Chem. 251, 5310–5314. 21. Citron, B. A., Davis, M. D., Milstien, S., Gutierrez, J., Mendel, D. B., Crabtree, G. R., and Kaufman, S. (1992) Proc. Natl. Acad. Sci. USA 89, 11891–11894. 22. Donlon, J., and Kaufman, S. (1977) Biochem. Biophys. Res. Commun. 78, 1011–1017. 23. Martinez, A., Knappskog, P., Olafsdottir, S., Doskeland, A. P., Eiken, H. G., Svebak, R. M., Bozzini, M. L., Apold, J., and Flatmark, T. (1995) Biochem. J. 306, 589–597.

AID

ARCH 9579

/

6b1f$$$$44

07-29-96 14:43:01

95

24. Hasegawa, H., and Kaufman, S. (1982) J. Biol. Chem. 257, 3084– 3089. 25. Ayling, J. E., and Helfand, G. D. (1975) in Chemistry and Biology of Pteridines (Pfleiderer, W., Ed.), pp. 305–319, de Gruyter, Berlin. 26. Døskeland, A. P., Døskeland, S. O., Ogreid, D., and Flatmark, T. (1984) J. Biol. Chem. 257, 11242–11248. 27. Phillips, R. S., and Kaufman, S. (1983) Trans. N. Y. Acad. Sci. 41, 87–96. 28. Døskeland, A. P., Schworer, C. M., Døskeland, S. O., Chrisman, T. D., Soderling, T. R., Corbin, J. D., and Flatmark, T. (1984) Eur. J. Biochem. 145, 31–37. 29. Smith, S. C., Kemp, B. E., McAdam, W. J., Mercer, J. E. B., and Cotton, R. G. (1984) J. Biol. Chem. 259, 11284–11289. 30. Kaufman, S. (1986) in Advances in Enzyme Regulation (Weber, G., Ed.), Vol. 25, pp. 37–64, Pergamon, Oxford/New York. 31. Yamashita, M., Minato, S., Arai, M., Kishida, Y., Nagatsu, T., and Umezawa, H. (1985) Biochem. Biophys. Res. Commun. 133, 202–207. 32. Woo, S. L. C., Gillam, S. S., and Woolf, L. I. (1974) Biochem. J. 139, 741–749. 33. Choo, K. H., Cotton, R. G. H., Danks, D. M., and Jennings, I. G. (1979) Biochem. J. 181, 285–294. 34. Abita, J. P., Parniak, M., and Kaufman, S. (1984) J. Biol. Chem. 259, 14560–14566. 35. Fisher, D. B., and Kaufman, S. (1973) J. Biol. Chem. 248, 4345– 4353. 36. Shiman, R. (1985) in Folates and Pterins (Blakley, R. L., and Benkovic, S. J., Eds.), Vol. 2, pp. 179–249, Wiley, New York. 37. Stegink, L. D., Filer, L. J., Jr., Brummel, M. C., Baker, G. L., Krause, W. L., Bell, E. F., and Ziegler, E. E. (1991) Am. J. Clin. Nutr. 53, 670–675.

arcalc

AP: Archives